1193. Comparison of the clinical outcomes of patients with IMP-type carbapenemase-producing carbapenem-resistant Enterobacteriaceae and carbapenem-sensitive Enterobacteriaceae in Japan.
Session: Poster Abstract Session: Healthcare Epidemiology: MDR-Gram Negative Infections
Friday, October 5, 2018
Room: S Poster Hall
Background: Carbapenem-resistant Enterobacteriaceae (CRE) infections are spreading worldwide and have become a global menace. Different types of carbapenemases contribute to carbapenem resistance. The outcome of patients with IMP-type carbapenemase-producing CRE (IMP-CRE) is not well known.

Methods: A matched case-control study from 1/2012 – 12/2016 was conducted at NCGM. All unique patients with IMP-CRE isolation were included and matched with carbapenem-sensitive Enterobacteriaceae (CSE) patients. Meropenem non-susceptible and/or ceftazidime-resistant Enterobacteriaceae, per CLSI criteria, were tested for metallo-b-lactamase production with further confirmation of blaIMP by PCR. Multivariate analyses were conducted for outcomes, adjusting for a propensity score predicting the likelihood of isolation of CRE versus CSE. The balance of each group was determined by standardized biases <0.25 for variables on baseline characteristics.

Results: In total, 192 patients (96 CRE, 96 CSE) were included (132 Enterobacter sp., 60 Klebsiella pneumoniae). Isolations sites were sputum (n = 76 [39.6%]), urine (n = 62 [32.3%]), blood (n = 22 [11.3%]), and wound (n = 14 [7.2%]). The median age of the patients was 75 years [IQR: 66-84], and 109 (56.8%) were male. Thirty-one (32.3%) patients with CRE and 55 (57.3%) patients with CSE developed infections. The others were considered as colonization. qSOFA was positive (≥ 2) in 7 patients with CRE infection and 9 with CSE infection. In bivariate analysis, mortality and length of hospital stay (LOS) after CRE/CSE isolation were similar between the 2 groups, even after stratification by bacterial species and infection/colonization. After controlling for the propensity score (Table), mortality and LOS remained similar between the 2 groups.

Conclusion: IMP-CRE might not contribute to the worsened clinical outcomes as compared to CSE. Further evaluations are needed for additional outcome parameters.

Table. Outcomes for isolation of IMP-CRE , n (%)






Multivariate analysis*

Bivariate analyses

Adjusted odds ratio (95% CI)

P value

P value

30-day mortality

15 (15.6)

17 (17.7)

0.85 (0.33-2.20)



Median length of stay (LOS) after the CRE/CSE isolation excluding death, days (IQR)





1.33 (0.84-2.13)



* Controlled for propensity score

Sho Saito, MD1, Kayoko Hayakawa, MD, PhD1, Masahiro Ishikane, MD, PhD1, Taichi Tajima, RN2, Nobuaki Matsunaga, MD, PhD, MPH3, Satoshi Ide, MD1, Noriko Kinoshita, MD1, Yoshiki Kusama, MD2, Yumiko Fujitomo, MD, PhD2, Takato Nakamoto, MD1, Yuta Toda, MD1 and Norio Ohmagari, MD, MSc, PhD1, (1)Disease Control and Prevention Center, National Center for Global Health and Medicine (NCGM), Tokyo, Japan, (2)AMR Clinical Reference Center, National Center for Global Health and Medicine, Tokyo, Japan, (3)AMR Clinical Reference Center, National Center for Global Health and Medicine Hospital, Tokyo, Japan


S. Saito, None

K. Hayakawa, None

M. Ishikane, None

T. Tajima, None

N. Matsunaga, None

S. Ide, None

N. Kinoshita, None

Y. Kusama, None

Y. Fujitomo, None

T. Nakamoto, None

Y. Toda, None

N. Ohmagari, None

Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 3rd with the exception of research findings presented at the IDWeek press conferences.